On the Adversarial Robustness of Graph Neural Networks with Graph Reduction

Kerui Wu, Ka-Ho Chow, Wenqi Wei, Lei Yu · 2024

Read On the Adversarial Robustness of Graph Neural Networks with Graph Reduction by Kerui Wu, Ka-Ho Chow, Wenqi Wei, Lei Yu (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-05883.txt.

Description

Kerui Wu and co-authors probe the safety of graph neural networks in a 2024 study. Shrinking huge graphs by merging nodes or dropping edges makes these models fast enough for big data, but whether shrinking opens attack holes was untested. The team runs poisoning attacks on many data sets and model types under ten reduction methods.

Dropping edges blunts some attacks yet barely dents others. Merging nodes does the opposite, amplifying attacks and sinking accuracy as graphs shrink further. The authors trace why each effect happens and test defensive models under reduction. They offer design guidance for fast graph systems that must also resist tampering.

Catalog record

CREATORKerui Wu, Ka-Ho Chow, Wenqi Wei, Lei Yu
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE58 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-05883

Browse the full library · About the Tower of Babel